Custom Search

Thursday, November 5, 2009

The "Long Tail" of WSN and the Smart Semiconductor Vendors (2)

Realizing the long tail of WSN will open your eyes and ears if you are on the user side who needs to press the button. For people like me, whose opinions, though, are interesting but most of the time irrelevant, catching the long tail at least make these industrial conferences a lot more entertaining, though sometime in retrospect. 

Often when a group of companies all touted one technology, the presenters of those long-tailed semiconductor vendors stood up and said something astonishing that made everybody awake.
Two years ago, I was in the probably first Bluetooth Low Energy conference, while many BT pure-players and supporters were really high, the guy from Nordic semiconductor claimed "proprietary solutions will just not go away" followed by his comparison that at certain points like power consumption and cost, proprietary solutions had its advantage. In the same conference, when people criticizing Zigbee and chanting BTLE would deliver, the guy from Texas Instruments calmly explained that Zigbee is infrastructure based and is more suitable than BTLE in addressing certain problems.

That was when Bluetooth Low Energy was still in vapor.
They can not say it too obviously in that kind of environment, but the message the ventors want you to get was "don't wait, come and buy our existing products."

Now the BTLE specification will be released before Christmas; chips will be in the market next year. The messages from the long-tailed companies have been changed too. On the surface, they become more aggressive. TI did demonstration in the SIG's first BTLE conference. Nordic has reported more often about its uBlue chips and progress. But they are playing a balancing game.

Sometimes the game were not really well played (and therefore entertaining!) For example, the folder to all attendees of BTLE SIG conference in Munich contained the Nordic newsletter, which like their website, mixes materials promoting BTLE, ANT and their proprietary solutions. The funny thing was how this newsletter could skip the eyes of the SIG, so that on the stage it was said BTLE will claim the death of other technologies; off the stage, these technologies seem to be able to live harmoniously together.

TI did not allow BT SIG the chance to make this kind of mistake. But if I presented two product sheets, one is from the ANT+ symposium early Oct, the other is from the Munich conference two weeks later. What do you think?

I mentioned that BT SIG may regret to have BTLE radio too similar to its existing competitive solutions. But this makes the semiconductor vendors, like Nordic and TI, happy. The benefit for them is one chip architecture can support different protocols. If you do not believe, you can correlate their relevant product press announcements and have a guess. TI is even going one more step further. They have made the chips of the two totally different radio technologies, 802,15,4 (or Zigbee, SimpliciTI etc.) and BTLE, share the same pin-out.

Don't be confused when you hear a company say this in one conference and say something else in another; or see them promote competing technologies. These vendors are just saying "come to buy our products, any one of them and up to you!"

If you are struggling comparing technologies and making up your mind. You need to be clear 1. there is no perfect solution 2. trade-off is required based on a thorough understanding of your application requirements. A long-tailed vendor may give you the better technical assistance.

1 comment:

  1. Excellent post. There is definitely not a one size fits all WSN protocol available today. Users should ignore the hype and look at their own requirements and existing solutions.